Tom Goldstein

This page is outdated. You will be redirected to my new web page at

I am a post-doctoral scholar in the Department of Electrical and Computer Engineering at Rice University, working in Rich Baraniuk’s  Digital Signal Processing group. I obtained my PhD in Applied Mathematics at UCLA in 2010 under the supervision of Stanley Osher, and completed a post-doctoral fellowship at Stanford University under Stephen Boyd. 

    My primary research interest is in numerical methods for optimization problems and elliptic PDE's. Most of my work has applications in image processing, especially for magnetic resonance imaging (MRI) and computed tomography (CT) technology. My current research involves large scale optimization for L1 regularized problems, and fast reconstruction methods for CT using algebraic models. 

   A common thread in my research is the development of fast algorithms for large scale problems - problems for which conventional ‘off the shelf’ techniques are not computationally tractable.  For many such problems, extremely efficient algorithms can be developed by exploiting the algebraic, geometric, or PDE structure of the problem.

    My research on optimization methods for image processing and computer vision has recently been featured by Thomson Reuters Sciencewatch as a ‘New Hot Paper in Computer Science.‘   More information on this work, and my interview with Thomson Reuters, can be found here.

Please visit the webpages for the various software tools I have developed:

Perfusion Imaging Toolkit


Split Bregman Methods




Applied Mathematics